Protracted Opioid Use Not Uncommon After Cardiothoracic Surgery

Preoperative prescriptions and higher doses predicted persistent use 3 to 6 months later, hinting at windows for prevention.

Protracted Opioid Use Not Uncommon After Cardiothoracic Surgery

One in eight patients who undergo cardiothoracic surgery continue to take opioids long after pain control is necessary, according to a new analysis of Medicare data from 2009 to 2015.

While some predictors of persistent use, like female sex, younger age, and existing pain disorders, aren’t modifiable, others are: opioids are more likely to become a problem for patients when prescribed before surgery or at higher doses. This, researchers say, draws attention to the fact that certain measures may help patients avoid opioid overuse, which can lead to subsequent addiction.

Prior studies have addressed opioid use after a variety of surgeries—orthopedic, gynecologic, dental, urologic, and others—but the current report provides a unique look specifically at cardiothoracic surgery, Alexander A. Brescia, MD (University of Michigan, Ann Arbor), and colleagues note in their paper published online August 21, 2019, ahead of print in the Annals of Thoracic Surgery.

To TCTMD, Brescia said the rate of persistent use seen here with cardiothoracic surgery—12.8% overall—is higher than the 3% to 8% previously shown for general surgery. “To be above 10% is notable,” he observed, adding that one bit of good news is that persistent use grew less prevalent over the course of the study.

Moreover, it’s important to make distinctions between persistent opioid use, dependence, and addiction, Brescia stressed. Persistent use puts patients at risk for opioid dependence; this dependence, in turn, “definitely confers morbidity and adverse effects on health but is not equivalent to addiction, which necessarily alters an individual’s life, whether personally, socially, or financially,” he said.

Vinod Thourani, MD (MedStar Heart and Vascular Institute, Washington, DC), commenting on the results for TCTMD, described the study as “very interesting, provocative, and timely.”

Thourani emphasized, though, that the procedures covered in this snapshot were very heterogeneous. For example, less than 1% were TAVR cases. “I think that we have to be careful in just relating this to cardiac-surgery patients compared to thoracic-surgery patients. We all know that a thoracotomy hurts more than a sternotomy incision does,” he said.

One big strength, he said, is that it identifies modifiable risk factors, specifically “the [timing] of prescriptions that we give, the doses we give them at, and how that really affects persistent drug abuse.”

Time and Dose Matter

The investigators tracked 24,549 Medicare patients (mean age 71 years; 38% women) who, according to claims data, were opioid naive from at least 1 year before until 30 days prior to cardiothoracic surgery. Procedures included isolated CABG (49.9%), isolated valve repair/replacement (16.0%), CABG plus valve intervention (12.6%), minimally-invasive lung resection (10.7%), open lung resection (9.3%), esophagectomy (0.8%), and TAVR (0.7%). All patients filled at least one opioid prescription between 30 days before surgery and 14 days after discharge; fully 6.9% of obtained their opioids prior to surgery.

In terms of factors that might increase the risk of opioid abuse, 47% of patients were current or previous tobacco users, 25% had anxiety or mood disorders, and 2% had drug or substance abuse disorders. Nearly half had a complication, the most common being arrhythmia and blood transfusion, each affecting 20% of patients. Hospital stays were longer for those who experienced a complication (mean 8.5 vs 6.6 days; P < 0.001) but their prescription sizes tended to be smaller (mean 353 vs 384 oral morphine equivalents [OME]).

New persistent opioid use, defined as continued prescription refills at 91 to 180 days after surgery, was seen in 12.8% of patients overall. The proportion declined yearly from 17.0% in 2009 to 7.1% in 2015 (P < 0.001).

On multivariable regression analysis, numerous factors predicted persistent use.

Predictors of New Persistent Opioid Use After Cardiothoracic Surgery


Adjusted OR

95% CI

Prescription Filled Before Surgery



Black Race



GI Complications



Disability the Reason for Medicare



Eligible for Both Medicare/Medicaid



Open Lung Resection






Drug/Substance Abuse



Female Sex



Tobacco Use




Additional predictors were higher predicted 10-year death risk (Charlson Comorbidity Index > 3) and the presence of pain disorders such as arthritis or back pain.

Importantly, there also was a link between prescription size and new persistent use. Patients given doses greater than the 75% percentile size (> 450 OME) were nearly twice as likely to continue on opioids than those below the 25th percentile ( 200 OME), at 19.6% versus 10.4%, respectively (P < 0.001).

Asked why opioid scripts might be given ahead of surgery, Brescia replied that he can’t know for sure from this data set, but a number of reasons are possible. “It could be logistical—just that at the preoperative appointment, some surgeons prefer to give the prescription then, to be used postoperatively—or it could be if the patient is in pain for some reason in that 30-day window before surgery,” he conjectured.

‘Tip of the Iceberg’

Thourani confirmed that awareness is growing about the potential for opioid misuse after surgery. One approach being tried is use of enhanced recovery after surgery (ERAS) protocols, part of which involves trying to decrease the amount of opioids unnecessarily given to patients.

“We believe that this decrease in opioids is better for patients while they’re recovering—ie, they get fewer small bowel issues [and] they’re a lot more lucid when they’re here, so they’re walking and breathing better,” Thourani explained. “We believe that minimizing opioids perioperatively and also postoperatively will not only decrease complications like pneumonia and ‘goofiness’ after surgery, it will also get them up and mobilized faster. We do believe this is an important thing, and this is on top of, obviously, the opioid addiction that can occur postoperatively.”

This latest report, he added, is “really the tip of the iceberg of the studies that we need to start doing as a surgical community” to address the opioid epidemic.

For Brescia, the next logical step in research is looking beyond the prescriptions themselves to assess how many of the prescribed pills patients actually take. “There really is not a one-size-fits-all X amount of pills that patients should get,” he explained, suggesting that what people require while in the hospital might help guide the amount of opioids they receive at discharge.

Caitlin E. Cox is News Editor of TCTMD and Associate Director, Editorial Content at the Cardiovascular Research Foundation. She produces the…

Read Full Bio
  • Brescia reports being supported by the National Research Service Award postdoctoral fellowship.
  • Thourani reports no relevant conflicts of interest.